The Philadelphia chromosome (Ph1), present in > or = 95% of chronic myelogenous leukemia (CML) patients, is a well-characterized translocation that results in a unique chimeric gene product (BCR/ABL) with transforming capability. Molecular methods utilizing the polymerase chain reaction (PCR) to detect BCR/ABL mRNA transcripts has been useful for detecting minimal residual disease (MRD) after treatment, as well as for establishing the diagnosis of CML. Amplification-based assays for the BCR/ABL transcript, however, have shown variable reproducibility and sensitivity. This variability may be largely due to technical differences and insufficient controls. In this report, we describe the development of a highly controlled, reproducible, and sensitive PCR assay to detect Ph1 that is well suited to clinical and research applications. A validation study of 82 samples was performed consisting of 25 dilutions of K562 cells (Ph1+) into normal cultured B cells, 26 pre- and post-transplant peripheral blood samples from CML patients, 16 peripheral blood samples for diagnosis of CML, and 15 peripheral blood samples from healthy individuals. RNA isolated from 3 to 5 million leukocytes was reverse transcribed (RT) and amplified by nested primer PCR. The products were characterized using agarose gel electrophoresis. Approximately 1000 Ph1-positive cells admixed with 10(6) normal cells were detectable after one round of amplification. In 60% of assays where one Ph1-positive cell was admixed with 10(6) normal cells, a BCR/ABL product was detectable after nested primer PCR. Specific measures to ensure accurate results in routine testing included (a) assessing RNA integrity and adequate cDNA preparation by detection of the constitutively expressed ABL mRNA, (b) monitoring sensitivity with the addition and detection of K562 RNA mixed with RNA from unknown samples (failure to detect the "spiked" K562 RNA indicates the presence of inhibitors or ribonucleases within the unknown RNA sample), (c) detection of nucleic acid contaminants by using negative controls in every assay, and (d) duplicate analysis of all samples and controls. Internally, this assay was 100% reproducible. Our results verify that nested primer RT-PCR is a fast, sensitive alternative to cytogenetic or Southern blot analysis for monitoring MRD after treatment and for diagnosis of CML. In addition, the highly controlled detection scheme presented here can be used as a general model for the development of other amplification-based detection assays.